At Tata Consultancy Services, I build enterprise AI infrastructure that connects Azure OpenAI, OpenAI, and Gemini through a unified FastAPI gateway supporting 100+ RPS.
I focus on making GenAI systems reliable, secure, and cost-efficient through prompt optimization, Redis semantic caching, asynchronous processing, and modular agent-based workflows.
I've built an LLM Gateway that reduced input tokens by approximately 30%, enabled parallel execution within rate limits, and enforced JWT authentication, Pydantic validation, and Pytest coverage. I also created a Figma-to-frontend automation pipeline that reduced design-to-implementation turnaround by 70%.
My work spans multimodal AI, medical symptom intelligence, vector search, and predictive healthcare models. I bring a backend-first approach to building production-ready AI applications using Python, FastAPI, LangGraph, PostgreSQL, and Redis.
